4 SEVENTH REPORT—1837- 
in May 1792; at Amboyna, in October of the same year; again at 
Van Diemen’s Land, in February 1793; and at Surabaya in 
Java, in 1794. With this last observation the published results 
terminate; there is no record of the vibrations having been re- 
peated on the return to France, for the purpose of testing the 
constancy of the magnetism of the needle, a step which subse- 
quent experience has shown to be most important. ‘The con- 
nexion of all the foreign stations with Europe is consequently 
imperfect; and the values of the intensity at those stations, re- 
latively to any standard value in Europe, could only be com- 
puted, subject to the uncertainty arising from the possibility 
of a change in the magnetic condition of the needle. The 
conclusion drawn by M. de Rossel, of the increase of the inten- 
sity in receding from the equatorial to the higher latitudes, was, 
however, fully borne out and substantiated, in regard to the 
southern hemisphere, by the observations at Van Diemen’s Land 
in 1792 and 1793, compared with the intermediate vibrations at 
Amboyna. These form a comparison complete in all respects, 
and to the certainty of which nothing is wanting. It is inde- 
pendent of any change the needle may have undergone before 
or afterwards ; the correspondence of the time of vibration at 
Van Diemen’s Land in May 1792 and February 1793, proving 
the needle to have been steady in that interval. ‘The increase 
in the intensity between Amboyna and Van Diemen’s Land was 
in the proportion of 1 to 1°67, a difference far too great to be 
attributed to any supposable errors or accidents of observation. 
It is this determination which unquestionably entitles Admiral 
de Rossel to the distinction which he has always enjoyed, of 
having been the first who ascertained that the magnetic inten- 
sity is different at different positions on the earth’s surface: al- 
though his observations were not published until after those of 
M. de Humboldt in 1798-1803, by which the same fact was 
more largely established. 
As M. de Rossel’s observations have not, I believe, been pub- 
lished in any English work, I have subjoined a table containing 
an abstract of all their essential particulars. 
